01 — CONTEXT: New Client Kit
Input definitions, validation rules, and what each mode requires.
Mode 1 Inputs — Day 0 Scaffold
All inputs are provided by Kathryn at engagement start.
| Input | Required | Example | Used For |
|---|---|---|---|
| Client preferred name | Yes | Ruben, Jenn | File naming, CLAUDE.md header, all document prefixes |
| Client full name | Yes | Ruben Cruz, Jenn Barr | CLAUDE.md header, README, master plan file name |
| Company name | Yes | Crulliance, TRG Consultants | CLAUDE.md header, README |
| Client slug | Yes | rc, jb | Repo name (aos-client-rc), used nowhere else |
| Repo name | Yes | aos-client-rc | GitHub repo, local path, README |
| Engagement type | Yes | Accounting Manager Operating System Implementation | CLAUDE.md header |
| Rate | Yes | $1,500/month (founding rate, 6-month commitment) | CLAUDE.md header |
| Cadence | Yes | Weekly Mondays | CLAUDE.md header, WORKFLOW.md rhythm table |
| Start date | Yes | February 9, 2026 | CLAUDE.md header |
| Session day(s) | Yes | Monday, Friday | WORKFLOW.md rhythm table, agenda step naming |
| Stakeholder sessions | No | Thursdays (Pooja — build session) | WORKFLOW.md rhythm table, additional processing days |
| Entry point | No | Diagnostic call, Referral | Reference data engagement section |
| GitHub org | Yes | kathryn-art | README GitHub section |
Validation Rules — Mode 1
- Preferred name is used for all file prefixes.
ruben-reference-data.md,ruben-constraint-brief-*.md,client-master-plan-ruben-cruz.html. Never use the slug for file naming. - Slug is lowercase, 2-3 characters, derived from client initials. Used only in the repo name:
aos-client-XX. - Repo name follows the pattern
aos-client-XXwhere XX is the slug. No exceptions. - Cadence determines WORKFLOW.md structure. A client with one weekly session gets a simpler rhythm table than one with multiple sessions.
- Rate format is free text. Include any qualifiers (founding rate, commitment period, etc.).
Mode 2 Inputs — Session 1 Population
| Input | Required | Source | Used For |
|---|---|---|---|
| Session transcript(s) | Yes | Google Drive → Relay | CPM processing, reference data population, master plan |
| JSON analysis file(s) | Yes | Relay | CPM processing, master plan |
| Constraint brief | No | Client (if process is set up) | CPM processing — but never process alone, wait for transcript |
| Client's stated vision/goal | No | Transcript, intake form, prior conversations | Direction anchor for CPM Mode 1 |
| Pre-AOS transcripts | No | Google Drive | Additional context for first CPM run |
Validation Rules — Mode 2
- At least one session transcript is required. Cannot populate reference data or run CPM without conversation data.
- JSON analysis accompanies every transcript. If missing, process with transcript only but note the gap.
- Constraint brief is never processed alone. Always wait for the session transcript. The brief is prep — not a standalone input.
- Vision/goal is the primary Direction anchor for the CPM. If the client has declared an annual goal, a "year of ___" statement, or a stated vision in any setting, this anchors the entire diagnostic.
Mode 3 Inputs — Maturity Upgrade
| Input | Required | Source | Used For |
|---|---|---|---|
| 3+ processed sessions | Yes | Existing CPMs and master plan | Pattern evidence for brand/skill decisions |
| Established naming patterns | Yes | Existing builds and documents | Brand kit — what terms this client uses |
| Brand preferences | No | Client feedback, approved builds | Brand kit — colors, voice, terminology |
| Recurring build types | No | Project plan, build history | Skill creation — which doc types repeat |
Validation Rules — Mode 3
- Do not create brand files before session 3. Need enough pattern evidence. Exception: client brings existing strong brand.
- Skills must reference vault frameworks, not duplicate them. A client-specific SOP skill extends the vault deployment cycle — it doesn't rewrite it.
- QC agents layer on vault QC, not replace it. A client brand QC adds brand-specific checks on top of the vault's general copy QC.
- Each Mode 3 action is independent. You can create a brand kit without creating skills. You can create skills without creating QC agents. Each stands alone.
Input Availability by Mode
| Input | Mode 1 | Mode 2 | Mode 3 |
|---|---|---|---|
| Client name/company/slug | Required | Available | Available |
| Engagement details (rate, cadence, type) | Required | Available | Available |
| Session transcript | — | Required | Available |
| JSON analysis | — | Required | Available |
| Constraint brief | — | Optional | Available |
| Client's stated vision/goal | — | Optional | Available |
| 3+ processed sessions | — | — | Required |
| Established naming patterns | — | — | Required |
| Brand preferences | — | — | Optional |
| Recurring build types | — | — | Optional |